Silicon Valley Hasn’t Revolutionized Warfare—Yet

But this was 2016, the building was the command post for all U.S. air operations in the Middle East, and the airmen were coordinating thousands of flights a day from dozens of bases—all by hand. The world’s most sophisticated military was using whiteboards to schedule its air traffic.

Eric Schmidt, the former chief executive of Google, was appalled. He was visiting the air base with Raj Shah, the head of Defense Innovation Unit Experimental (or DIUx), an at-the-time new Pentagon office tasked with integrating commercial technologies into U.S. warfighting. Schmidt told Shah, “this is the most egregious misuse of IT that I’ve ever seen. You guys need to fix this—and right away.” Shah took note, and DIUx subsequently built an app that automated the process of scheduling aircraft refueling, saving 25 million gallons of jet fuel annually.

This story is one of many similar examples of technological dysfunction and reform recounted in Unit X: How the Pentagon and Silicon Valley Are Transforming the Future of War. Written by Shah and Christopher Kirchhoff, who together led the unit, the book tells the story of DIU, launched in 2015 by then-U.S. Defense Secretary Ash Carter to help the Pentagon “drill holes in the wall” between the Department of Defense and Silicon Valley, in Carter’s words. (The unit dropped the “x” in 2018.)

Unit X is a readable and insightful, if at times hyperbolic, account of how cutting-edge commercial innovation is shaping warfare and how U.S. bureaucracies have struggled—and, more rarely, succeeded—in their efforts to keep up.

This symbiotic relationship changed with the end of the Cold War. For decades, U.S. technological and industrial innovation had largely been driven by federal spending; in the 1990s, it became increasingly commercialized and globalized. As Carter noticed as early as 2001, “tomorrow’s defense innovations will largely be derivatives of technology developed and marketed by commercial companies for commercial motives.” In 1960, U.S. defense spending accounted for around 36 percent of global research and development. By 2019, that number was just over 3 percent. That year, Amazon, Apple, Google, and Microsoft each had a market capitalization larger than that of the entire U.S. defense industry.

The Pentagon struggled to adapt. As Shah and Kirchhoff emphasize, it remained reliant on a handful of giant defense contractors known as the “primes”—companies such as Raytheon, Lockheed Martin, and Northrop Grumman—that possessed the elaborate auditing, accounting, and lobbying systems necessary to navigate the Pentagon’s labyrinthine acquisitions process. Trained to protect taxpayer money, Pentagon procurement officials were inclined to avoid risk, making them a poor fit to acquire products developed by venture capital firms and tech companies whose business models were premised on making bets, moving fast, and iterating rapidly. Under federal acquisition rules, large contracts take 18 to 24 months to finalize. Start-ups couldn’t wait that long to close a deal.

The obstacles were cultural, too. By the 2010s, most Silicon Valley engineers had grown up watching the invasion of Iraq; they had little desire to design products to help the military kill or the intelligence community spy. Edward Snowden’s disclosures that the National Security Agency was collecting tech companies’ data intensified this mistrust: When Carter visited Silicon Valley in 2015—the first defense secretary to do so in two decades—Google refused to let him enter its campus.

As a result of these economic and cultural trends, Shah and Kirchhoff write, “the two-decade explosion of consumer electronics that began in the 2000s somehow became invisible to the Pentagon.” As late as 2016, while civilians used smartphones to call taxis or pay bills, many U.S. forces were “running software programs older than the officers using them.” This software was often designed by legacy defense contractors—“a folly akin to hiring Microsoft to build an aircraft carrier,” the pair write.

A pilot puts on his helmet in the cockpit of an F-35 fighter jet for a training mission at Hill Air Force Base in Ogden, Utah, on March 15, 2017.George Frey/Getty Images

And while Washington spent billions on weapon systems such as F-35 fighter jets—which took so long to develop that by the time they became operational, their processors were slower than the ones in most Americans’ pockets—U.S. adversaries, from the Houthis to the People’s Liberation Army, were finding creative ways to turn off-the-shelf commercial products into military technologies for a fraction of the cost and time.

This was the problem that Carter launched DIU to solve. Shah and Kirchhoff recount how DIU officials became experts in leveraging obscure legal authorities such as the Other Transactions Authority (OTA), which allowed them to speed up contracts by buying technology at scale from a company as soon as a successful pilot project was complete, instead of making the company go through more rounds of bureaucratic competition. (The OTA process has now accounted for more than $70 billion of Pentagon spending.) DIU employees also tried to bridge cultural divides between bureaucrats and coders, jettisoning their suits and ties for jeans and untucked shirts.

Shah and Kirchhoff’s account is particularly interesting on the opposition and mistrust that DIU faced within Washington. The unit triggered what the authors describe as antibodies from the bureaucracy’s immune system. They recount how, for example, in their first week on the job, two congressional staffers on the House Appropriations Committee tried to scrap DIU’s entire $30 million budget. One of the staffers objected to DIU because she worked for an Indiana congressman, and DIU’s money was primarily flowing to California—unsurprisingly, given its importance to the U.S. technology ecosystem. The other staffer was upset that Carter had once refused his request to use an Air Force Gulfstream to take a congressional delegation overseas.

At almost every stage of its growth, DIU would have to overcome inertia. Shah and Kirchhoff indulge in some personal score-settling, and parts of the book are enough to put anyone off ever working in government. But they bring a deep understanding of both Silicon Valley buzzwords and Pentagon acronyms, and their account is substantively interesting on the real difficulties inherent in pushing change through large, cumbersome bureaucracies.

A Ukrainian soldier uses the Starlink system during military exercises in the Chernihiv region, Ukraine in June 2023.Maxym Marusenko/NurPhoto via Getty Images

No one would dispute that these technologies are useful. But whether they are as revolutionary as some of Silicon Valley’s biggest boosters maintain is more contested. “The tech bros aren’t helping us too much in Ukraine,” William LaPlante, the Pentagon’s top acquisition official, said at a defense conference almost a year into the conflict. “It’s hardcore production of really serious weaponry” that “matters,” he said. “We’re not fighting [in] Ukraine with Silicon Valley right now, even though they’re going to try to take credit for it.”

For all the technological breakthroughs of recent years, LaPlante is right that much of the fighting in Ukraine resembles conflicts from the past century. Conventional artillery and tanks remain crucial for both sides. Rather than heralding a long awaited revolution in military affairs, the war represents what the analyst Stephen Biddle calls a “marginal extension of long-standing trends and relationships between technology and human adaptation.”

None of this makes DIU’s work less valuable: Stultifying bureaucracy, slow acquisition processes, and excessive risk aversion hinder the evolutionary adaptation necessary to maintain Washington’s military edge. And Shah and Kirchhoff are correct when they argue that the U.S. military needs to reassess its balance between quality and quantity: Relying on small numbers of exquisite and exorbitantly priced military systems is unlikely to be the best approach to a strategic environment in which the United States cannot declare by fiat that it will only wage short, non-attritional wars.

Still, the realities of modern warfare suggest that the influence of new technologies is, for now at least, somewhat less transformative than Silicon Valley’s most zealous advocates suggest. Ultimately, the scale of DIU’s impact on the Pentagon remains unclear. For all the important but discrete successes that the authors point to—the rapid delivery of an electric air taxi for the Air Force, a counter-unmanned aerial system used by special operations forces—it has produced fewer results at scale and has not transformed the Pentagon’s broader acquisitions process.

Last year, venture-backed companies won less than 1 percent of all Pentagon contracts; the handful of giant prime contractors continue to receive vastly more funds than the Silicon Valley-inspired defense ecosystem. And the Pentagon still generally requires years of planning and congressional deliberation before it buys products from start-ups in quantities significant enough to keep their businesses going, to the ongoing frustration of investors.

But judged against a more modest set of goals, DIU has succeeded. With an annual budget of less than one F-35 for most of its existence, it has shown that commercial technology is viable in military missions. It has demonstrated that the Department of Defense can be a realistic customer for start-ups. And it has helped bridge the cultural gap between Silicon Valley and the military.

In the past four years, at least $125 billion of venture capital has poured into start-ups that build defense technology, up from $43 billion in the previous four years. Dozens of former national security officials are now working in defense-Sultra1news venture capital or private equity as executives or advisors. And an expanding group of Silicon Valley companies has started speaking the language of national security. As the investor Marc Andreessen wrote in 2019, “I believe in the United States of America. I believe in a strong national defense. And I believe in Anduril,” the defense technology company that, according to its founder, Palmer Luckey, pledged on the first page of its first pitch deck that it would “save western civilization … as we make tens and tens of billions of dollars a year.”

In other words, the vibes have shifted: Out of some mixture of financial self-interest and genuine ideology, Silicon Valley is enlisting itself once again in the United States’ great-power struggles. It’s a picture that would look increasingly familiar to the men who built up Silicon Valley in the 1950s.